PROJECT SUMMARY/ABSTRACT Glioblastoma (GBM) is one of the most lethal types of cancers, with the life expectancy of patients only modestly improved over the past decades, making it essential to develop advanced therapy for efficacious tumor elimination. Immunotherapy, especially adoptive transfer of T cells engineered with chimeric antigen receptors (CARs), has shown clinical promise in some hematological malignancies, yet its effect against GBM remains questionable, attributable to the fact that therapeutic T cells targeting GBM have not yet been demonstrated in vivo persistence, and that the heterogeneity of GBM tumors allows for escape from treatments. The research proposed here aims to address the unmet clinical need for effective GBM therapy, with the approaches split into two aspects. First, in the F99 phase of the award, I will enhance the antitumor potency of CAR-T cells against GBM by modifying the selection of specific T cell subsets and the incorporation of appropriate costimulatory signals. My work has already demonstrated (completed dissertation research) that CD4+ CAR-T cells outperform CD8+ cells, mediating persistent and long-term antitumor effects. Meanwhile, I will further investigate the functional impact of the CD27-CD70 interaction between T cells and tumor cells, on CAR-T cell exhaustion and persistence. GBM-targeting CAR-T cells will be enriched for CD27+ or CD27- subsets, or modified to constitutively express the CD27 molecule, followed by functional and phenotypic evaluation of their long-term antitumor activity and exhaustion status. These studies will lead to refinement of CAR-T cell manufacturing, especially for therapies against GBM, generating more potent and persistent products. During the K00 phase of the award, I will investigate the role of GBM stem cells (GSCs) in mediating escape from targeted immunotherapy. More specifically, I will test the hypothesis that heterogeneous GSC subtypes are able to interconvert with each other in response to therapy, and subsequently lead to tumor relapse that is no longer targetable by previous treatments. Experimentally, I will isolate GSC subsets with classical/proneural (CL/PN) and mesenchymal (MES) phenotypes. Using the tumor heterogeneity model of GSC-derived organoids as well as orthotopic tumor xenografts, I will examine GSC subsets for their evolution during the time course of immunotherapies that target specific GBM subtypes. Further, using cell barcoding technology, I will test the existence of interconversion between CL/PN and MES GSCs, and evaluate its potential effect in mediating GBM tumorgenesis and escape from therapies. Together, the proposed research holds translational potential for developing effective anti-GBM strategies with the combination of long-lasting immunotherapy and the specific targeting of certain GSC subgroups. Therefore, the goals of this application well fit the mission of the National Cancer Institute.